Falls in old age can have serious consequences. The impact on health and social care is growing as the older population increases, but there are few data on falling amongst the “oldest old”. This study aimed to provide much needed information on this fastest growing section of the population: the epidemiology of falls and their consequences, the prevalence of potential risk factors and their predictive value in extreme old age. This study added a special investigation of falling in advanced old age to the 2002–2003 interviews of 110 over-90-year-olds from the Cambridge City over-75s Cohort, a population-based longitudinal study of ageing. The survey (90 women, 20 men) comprised a standardised nurse-administered questionnaire with cognitive assessment, quantitative heel ultrasound scans and functional performance measures: Timed Unsupported Stand, Short Physical Performance Battery (standing balance tests, gait speed and chair rising),180° turn, functional reach and hand grip strength. Data collection also included a year’s prospective monitoring of falls using a combination of weekly calendars and telephone follow-up, with reports from participants themselves and proxy informants. The study’s description of a representative population aged over 90 is valuable to service planners preparing for demographic change, revealing high levels of many fall risk factors. Detailed characterisation of functional status showed close agreement between reported disability levels and performance measures. This first population-based survey of skeletal fragility in the tenth decade found quantitative ultrasound measures markedly lower than in previous studies with younger old people. Skeletal fragility reflected weight-bearing functional test performance and reported current or past mobility. This first prospective study of falling amongst people aged over 90 in a representative population-based sample found falls are even more common than previously reported for very old people a decade younger: 60% fell at least once during follow-up, 45% more than once. Incidence was 277 falls / 100 person-years. The extent to which falls in advanced old age lead to serious consequences – both immediate and longerterm – has not previously been reported. In one year’s follow-up 54% of fall reports described the participant as being found on the floor. 82% of falls occurred alone, 80% of those who fell were unable to get up after at least one fall, and 30% suffered long lies of an hour or more. Four out of five times when someone fell alone and could not get up they did not use available alarms to call help. More than half the falls reported to the study, and three in ten of the falls resulting in any injury, had not been reported to any health care professional. Findings also showed high levels of injuries (38% of falls but 68% of fallers) including fractures – one man and 1 in 8 of the women who fell. One in three people had at least one hospital admission, 2/3 of them at least partly due to falling, 2/3 of these directly prompted by a fall. Mean total length of stay of fallers was 6 times that of non-fallers. 1 in 7 of those not already living in long-term care had moved into homes within a year, 80% of these prompted at least in part by falling. Falls, adverse consequences and skeletal fragility shared a pattern of strong associations with several key risk factors, particularly impaired mobility and characteristics typical of frailty. Fracture risk factors were also associated with skeletal ultrasound measures. Functional tests added no predictive value to reported clinical risk factors. The implications of this research for policy and practice are fully discussed in relation to the current developing situation and future projections, setting these novel findings in the context of existing knowledge summarised in an extensive literature review.
